Utterances are records of the individual turns in a conversation, attributed to individuals in that conversation. After adding utterances to a conversation, analysis for each is enqueued, and you receive a notification when the analysis is complete. Note that the final analysis of a conversation -- that is, the analysis performed once a conversation has ended -- includes detail that the analysis of individual Create Utterance responses do not include, as we do not know the full duration of the conversation.

It is acceptable to post transcripts after a conversation has ended; the analysis returned for an ended conversation will include the full analysis, assuming that the full scope of the conversation is known.

The Utterance Object

field

type

description

time_start

Javascript timestamp (ms)

The start time of the utterance.

time_end

Javascript timestamp (ms)

The end time of the utterance.

content

string

The content spoken in this utterance.

individual_src_id

string

The src_id of the Individual who spoke this utterance.